Reciprocal brain–behavior associations in gambling disorder revealed by theta cordance and directional predictive modeling

Background Gambling disorder (GD) is a behavioral addiction characterized by impaired impulse control and fronto-limbic cortical dysfunction. EEG theta cordance, which integrates absolute and relative power to index regional cerebral perfusion and metabolism, has not previously been combined with directional predictive modeling in behavioral addiction research. Objective To compare regional resting-state theta cordance between GD patients and healthy controls (HCs), and to examine directional predictive associations between cortical cordance and clinical severity indices. Methods In this retrospective observational study, 28 male patients with DSM-5 GD and 29 male HCs underwent 21-channel resting-state EEG (eyes-closed, 3-min artifact-free epochs; 4–8 Hz). Theta cordance was computed using the Leuchter algorithm and averaged into eight regions of interest. Clinical measures included the South Oaks Gambling Screen (SOGS), Beck Depression Inventory-II, Beck Anxiety Inventory, and illness duration. Group differences were tested with independent-samples t -tests (with Bayesian complements) and Pearson correlations; an exploratory vector autoregressive (VAR)-based directional predictive framework was applied to cross-sectional observations. Results GD patients showed significantly lower central (0.03 ± 1.18 vs. 1.25 ± 1.23; t (55) = −3.79, p d = −1.00) and higher left parietal cordance ( t (55) = 2.12, p = 0.039, d = 0.56), with trend-level occipital ( p = 0.058) and left temporal ( p = 0.067) differences. No cordance–clinical correlations were significant (all p > 0.10). VAR-based analysis identified asymmetric predictive associations: prefrontal ( F = 4.60, p = 0.042) and frontocentral cordance ( F = 6.65, p = 0.017) with SOGS; central cordance with illness duration ( F = 5.77, p = 0.024) and right temporal cordance ( F = 5.53, p = 0.027); and reciprocal clinical-to-cortical associations (SOGS→central, F = 4.82, p = 0.038; duration→right temporal, F = 4.53, p = 0.044). HCs showed temporo-occipital directional patterns without clinical correlates. Conclusion Cortical theta cordance and clinical severity in GD appear reciprocally associated within an exploratory directional framework. Findings are preliminary given the modest, male-only, single-site sample and cross-sectional design, and require longitudinal confirmation.